Vol. 623: 161–174, 2019 MARINE ECOLOGY PROGRESS SERIES Published July 30 https://doi.org/10.3354/meps13024 Mar Ecol Prog Ser OPENPEN ACCESSCCESS Light comfort zone in a mesopelagic fish emerges from adaptive behaviour along a latitudinal gradient Tom J. Langbehn1,*, Dag L. Aksnes1, Stein Kaartvedt2, Øyvind Fiksen1, Christian Jørgensen1 1Department of Biological Sciences, University of Bergen, Bergen 5020, Norway 2Department of Biosciences, University of Oslo, Oslo 0316, Norway ABSTRACT: Throughout the oceans, small fish and other micronekton migrate between daytime depths of several hundred meters and near-surface waters at night. These diel vertical migrations of mesopelagic organisms structure pelagic ecosystems through trophic interactions, and are a key element in the biological carbon pump. However, depth distributions and migration ampli- tude vary greatly. Suggested proximate causes of the migration such as oxygen, temperature, and light often correlate and therefore the causal underpinnings have remained unclear. Using meso- pelagic fishes and the Norwegian Sea as a study system, we developed a dynamic state variable model that finds optimal migration patterns that we validate with acoustic observations along a latitudinal gradient. The model describes predation risk and bioenergetics, and maximizes ex - pected energy surplus, a proxy for Darwinian fitness. The model allows us to disentangle the driv- ers of migration and make predictions about depth distribution and related fitness consequences along a latitudinal trajectory with strong gradients in environmental drivers and vertical distribu- tion of scattering layers. We show that the model-predicted vertical migration of mesopelagic fishes matches that observed along this transect. For most situations, modelled mesopelagic fish behaviour can be well described by a light comfort zone near identical to that derived from obser- vations. By selectively keeping light or temperature constant, the model reveals that temperature, in comparison with light, has little effect on depth distribution. We find that water clarity, which limits how deeply light can penetrate into the ocean, structures daytime depths, while surface light at night controlled the depth of nocturnal ascents. KEY WORDS: Antipredation window · Deep scattering layer · Photoperiodic constraint hypothesis · High-latitude · Myctophidae · Twilight zone · Norwegian Sea · Benthosema glaciale 1. INTRODUCTION son et al. 2013, Aumont et al. 2018). Re vised esti- mates suggest that the mesopelagic zone, from ca. The mesopelagic is the daytime twilight zone in 200− 1000 m depth, harbours the majority of the the world oceans. The ecology here is a game of ‘hide global fish biomass (Irigoien et al. 2014). These di - and seek’, where organisms with limited swimming verse midwater assemblages make up the deep ability, e.g. micronekton such as small fish, crusta - scattering layer (DSL), which is distributed across all ceans, and siphonophores, try to avoid encounters major oceans from the tropics to sub-Arctic latitudes with larger predators, mainly bigger fish and squid. (Gjøsaeter & Kawaguchi 1980, Sutton et al. 2017). Mesopelagic micronekton are an important link in While scattering layers may occur in Arctic waters, the pelagic food web (Dagorn et al. 2000, Connan et they are mostly at tributed to advection from lower al. 2007, Naito et al. 2013) and might play a crucial latitudes (Gjøsæter et al. 2017, Knutsen et al. 2017, role in mediating climate change effects through the Saunders et al. 2017), and mesopelagic biomass sequestration of carbon into the deep ocean (Davi- strongly declines towards the poles (Kristoffersen & © The authors 2019. Open Access under Creative Commons by *Corresponding author: [email protected] Attribution Licence. Use, distribution and reproduction are un - restricted. Authors and original publication must be credited. Publisher: Inter-Research · www.int-res.com 162 Mar Ecol Prog Ser 623: 161–174, 2019 Sal va nes 1998, Siegelman-Charbit & Planque 2016, pelagic fish being out of reach for many air-breathing Escobar-Flores et al. 2018). predators (Horning & Trillmich 1999, Benoit-Bird et In synchrony with the day−night cycle, the DSL al. 2009a). rises and falls in the water column (Bianchi & Mislan When corrected for light attenuation, DSLs along a 2016), often spanning several hundred meters in circumglobal transect were found within a narrow depth (Klevjer et al. 2016). About half of all meso- range of light intensities (Aksnes et al. 2017), also pelagic micronekton, possibly more than 5000 mil- referred to as optical depth or light comfort zone lion tons, takes part in this diel vertical migration (Røstad et al. 2016a,b). With better vision at dim light (Irigoien et al. 2014, Klevjer et al. 2016). This makes and operating at smaller spatial scales, mesopelagic the daily vertical movements of mesopelagic organ- fish might be able to exploit light comfort zones as an isms the largest migration of living biomass on the antipredation window, so that their search efficiency planet (Hays 2003). is maximized relative to the mortality risk from visu- The depth range of the DSL migration varies greatly ally feeding piscivores (Clark & Levy 1988, Scheuer- (Bianchi et al. 2013, Klevjer et al. 2016). The daytime ell & Schindler 2003). A proximate light comfort zone sound scattering layer is closer to the surface in the mechanism has been suggested to control DSL depth northern Indian Ocean, the central Atlantic Ocean, in the Norwegian Sea, but the potential role of and the North and eastern tropical Pacific, than in the temperature in governing DSL distribution patterns subtropical gyres, the central Indian Ocean, the west- could not be excluded (Norheim et al. 2016). ern tropical and southeast Pacific, and the Pacific sec- Some fish species may use vertical migration as a tor of the Southern Ocean (Bianchi et al. 2013, Klevjer strategy to exploit thermal gradients to optimize et al. 2016). Some studies suggest these basin-scale energy budgets. Feeding in warmer, food-rich sur- patterns can be explained by the distribution of dis- face waters but resting and digesting at colder depth solved oxygen, with shallower migration depth in can help conserve energy because metabolic costs hypoxic areas (Koslow et al. 2011, Bianchi et al. 2013, in crease at higher temperatures (Rosland & Giske Netburn & Koslow 2015). Mesopelagic organisms 1994, Sims et al. 2006). However, warmer waters also forming the DSL, however, did not avoid oxygen- promote digestion, allowing fish to feed more and depleted waters. In contrast, in areas with oxygen therefore grow faster (Wurtsbaugh & Neverman minimum zones, these mesopelagic organisms ap- 1988). Such trade-offs can result in diverse, state- peared at hypoxic or even anoxic depths (Tont 1976, dependent migration and foraging strategies. Bianchi et al. 2013, Klevjer et al. 2016, Aksnes et al. The deep scattering layer depth also varies with 2017), suggesting no direct causation between their latitude (Beklemishev 1981). There are strong latitu- depth distribution and dissolved oxygen. dinal gradients not only in temperature but also in Light penetration has been pointed to as a parsimo- seasonality of surface light and day length. Moving nious, mechanistic, and universal explanation for DSL poleward, the distinct day−night light cycle gradu- depth distribution. Light was among the first drivers ally changes into a seasonal light regime with little suggested to structure DSL depth (Kampa 1971, Dick- variation in light intensities over the diel cycle. Dur- son 1972, Tont 1976). Since then, various lines of evi- ing the Arctic summer, the sun never sets, and it may dence have converged on light being a first-order not rise for months at a time during the polar night. driver of DSL behaviour and therefore the most defin- These patterns have strong implications for the verti- ing environmental factor for mesopelagic ecosystem cal extent of the twilight zone (Kaartvedt et al. 2019). structure (reviewed by Kaartvedt et al. 2019). For ex- Consistent with the assumption that light penetration ample, DSL species change their depth distribution in structures vertical migration depth (Aksnes et al. response to the lunar cycle (Benoit-Bird et al. 2009b, 2017), the DSL was deeper during summer at higher Wang et al. 2014, Prihartato et al. 2016) but also to latitudes, where observed migration was of smaller short-term perturbations in light level from weather amplitude and not as distinct (Sobolevsky et al. 1996, (Barham 1957, Kaartvedt et al. 2017), eclipses (Backus Norheim et al. 2016). et al. 1965, Tont & Wick 1973, Kampa 1975), and vari- In this study, we aim to quantify the role of light ver- ations in waters of changing transparency (Abookire sus other drivers in structuring the depth distribution et al. 2002, Norheim et al. 2016). Full moon, cloudless of mesopelagic fishes, with an emphasis on latitudinal skies, and clear water cause deeper nocturnal scatter- gradients and the associated change in light regime ing layers, with ramifications through the food chain towards high latitudes. Previous studies, most of them because of reduced predation on zooplankton in the empirical, have described migration patterns and es- upper waters (Hernández-León et al. 2010) and meso- tablished correlations with potentially relevant driv- Langbehn et al.: Emergent light comfort zone in mesopelagic fishes 163 ers, pointing to light as the most dominant on a global wegian Sea (63.77°−68.80° N), from the Norwegian scale (Aksnes et al. 2017). Here, we move beyond cor- shelf break in the southeast, across the Arctic Circle, relational inference from observational data and de- to the Icelandic plateau in the northwest (Fig. 2A). velop a theoretical and evolutionary model, built from Water temperature, water clarity, and surface light in- simple, well-understood, and quantifiable mechanisms tensities all showed marked south-to-north gradients of visual encounters between prey and their predators (Fig. 2B−D), and the timing was 1 wk before the onset (e.g. Eggers 1977, Aksnes & Giske 1993) and tempera- of the midnight sun period in the northernmost part of ture-dependent bioenergetics (e.g.